2. Defining the market
Face - Contents - Bottom - Previous/Next "A Powerful Competition Policy" 2. Defining the market2.1 The sub-markets of the electricity sectorThe power market can be divided into a physical market and a financial market. These two markets can be divided further into several sub-markets. We will describe briefly the various markets. Note, however, that these sub-markets do not necessarily constitute relevant markets as defined by competition authorities. It is common to divide the power market into three different vertical levels. One level is transmission and distribution of power through the grid. A second level is the wholesale market where producers, suppliers, large industrial enterprises and other large units buy and sell electricity, and a third level is the retail market where power is sold from retail power companies to end users such as households and firms. The transmission and distribution of electric power is usually regarded as a natural monopoly and therefore regulated in all the Nordic countries. There is no competition on this level of the power market. Since this report is aimed at preserving and developing competition in the market for electricity in the Nordic region, we will not focus on the transmission and distribution of power. The division between the wholesale market and the end users market is sketched in the figure below.
Electricity is sold to most end users through power suppliers (i.e. retail power companies). Some of these companies also produce power, while others buy all the needed power on the wholesale market. In the retail markets there are barriers restricting purchase of power cross borders. The main reasons are differences in each countries regulatory regime, taxes and prices, which make cross border trade difficult. As a result, retail power companies are established in the country where they sell power. The retail markets do therefore also fall outside the scope of this report, since they contain few competitive issues relevant for the Nordic electricity market. Based on the aforesaid we will limit our focus in this report to the wholesale market. 2.2 Definition of the relevant marketIn competition analysis the aim is usually to determine whether one or more companies have or can acquire market power. To do this, competition authorities usually define the relevant market where the potential market power is or could be abused. The market definition is a tool used to identify and define the boundaries of competition between firms. In the U.S. horizontal merger guidelines the relevant market is defined as follows:
This test is also called the "SSNIP test" (Small but Significant and Nontransitory Increase in Price). The SSNIP test defines a relevant market as the narrowest collection of products over which a hypothetical monopolist would find it profitable to make a small but significant and non-transitory increase in price (SSNIP). The U.S. market definition focuses solely on demand substitution factors--i.e., possible consumer responses. Supply substitution factors – i.e. possible production responses – are considered elsewhere in the Guidelines in the identification of firms that participate in the relevant market and the analysis of entry. The European Commission's definition of a relevant market separates between the relevant product market and the relevant geographic market:
At the outset the two definitions differ somewhat. The EU definition of a relevant product market seems to include all substitutable products. Seemingly, this definition will include more products than what is necessary for the SSNIP test to be satisfied (it will be economically profitable to increase price even if some substitution to other products will take place, i.e. all substitutable products need not be included in the relevant market). Furthermore, the EU definition of a geographic market focuses on other aspects than whether demand substitution (or other restraints on market power) will hinder the SSNIP test from being satisfied. However, in the Commission Notice on the definition of the relevant market (which reflects the jurisprudence of the Court of Justice and the Court of First Instance) the Commission interprets the definitions. The clarification makes clear that the SSNIP is applied both to the product and to the geographic market definition. The Notice points out that firms are subject to three main sources of competitive constraints: demand substitutability, supply substitutability and potential competition. From an economic point of view, for the definition of the relevant market, demand substitution constitutes the most immediate and effective disciplinary force on the suppliers of a given product, in particular in relation to their pricing decisions. The competitive constraints arising from supply side substitutability and potential competition are in general less immediate . According to the Notice, supply-side substitutability may also be taken into account when defining markets in those situations in which its effects are equivalent to those of demand substitution in terms of effectiveness and immediacy. This requires that suppliers are able to switch production to the relevant products and market them in the short term without incurring significant additional costs or risks in response to small and permanent changes in relative prices. The same reasoning applies to geographic areas. Thus, unlike the U.S. Guidelines the EU Commission Notice may take supply substitution into account when delineating markets. However, the U.S. Guidelines makes clear that supply substitution may be considered when identifying firms that participate in the relevant market. Firms that would quickly and easily substitute supply in response to an attempted exercise of market power are considered to be competitors in the same market and are assigned market shares even though they do not currently sell. In practise it seems likely that both the EU and the U.S. approach is capable of identifying supply substitution competition in a meaningful way, which will produce similar conclusions. [5] The hypothetical price increase is usually assumed to be 5% or 10%. The principal procedure to delineate relevant markets is to ask whether a hypothetical monopolist is able to profitably make such a price increase. If not, the market should be expanded by including the best substitute outside its borders, either in the product space or in the geographic space. By expanding the market in this way until the price increase becomes profitable, the relevant market will be delineated as the smallest group of products and the smallest geographic area in which the SSNIP test is satisfied. It is not clear what shall be meant by "non-transitory". This is of some importance since demand is likely to be more elastic in the long than in the short run. The U.S. Guidelines considers supply substitution within the framework of one year. Copenhagen Economics (2002) considers the SSNIP test within a time period not shorter than twelve months. The SSNIP test is particularly applicable to markets where the issue is whether a concentration will increase the price level in the market. It is less applicable as a tool to analyse whether a firm already possess market power. More on this subject can be found, for instance, in a paper from the Office of Fair Trading (2001). [6] 2.3 The relevant product marketThe European Commission has traditionally defined the product market as the sale of electricity delivered through high-voltage network. There are no close substitutes for electricity. In the short run (on an hourly basis) demand is very inelastic at current prices. The price sensitivity of demand is low also in the longer run - during a season or some months. This means that it will be possible for a hypothetical monopolist to impose a small but significant and non-transitory increase in price. [7] We therefore ascertain that wholesale of electricity is a separate product market. However, we have to investigate whether there are smaller relevant markets within the wholesale market (confer that the relevant market is the smallest market in which it is possible to exert market power). In particular we are interested in knowing if electricity sold through bilateral contracts belong to the same relevant product market as the sale of electricity at the Nord Pool elspot market. Examining this problem using the SSNIP test, we ask if a hypothetical monopolist in the bilateral market (or the spot market) would be able impose a small but significant and non-transitory increase in the electricity price without loosing profit due to loss of sales on the spot market (or the bilateral market). Due to the range of different bilateral contracts it is hard to be completely affirmative, but the most probable answer to the above-mentioned question would be that the hypothetical monopolist would not have the incentive to increase the price on bilateral contracts. This is due to two constraining effects of such a price increase. Firstly, buyers that have entered into bilateral contracts will to some extent switch to the spot market as a response to a hypothetical price increase. And buyers operating on the spot market will be able to switch to bilateral contracts. In the cases concerning Statkraft's acquisition of Agder Energi [8] and Trondheim Energiverk the Norwegian Competition Authority concluded that the bilateral market and the elspot market are close substitutes, both for suppliers and customers, and therefore part of the same relevant product market. An important argument was the fact that many bilateral contracts contain some reference to the system price. The decisions were appealed to the Norwegian Ministry of Labour and Government Administration who handled the cases in the second instance. The Ministry supported these findings. The Danish competition authority confirmed this reasoning in a recent case concerning abuse of dominant position by two Danish market participants. [9] The Danish competition authority argued that a bilateral contract in essence comprises a spot market element and a financial element. The Nord Pool spot market governs pricing of the spot element whereas the financial element belongs to at completely different market – the market for risk. At this market the "good" traded is risk and not electricity. (Risk is also traded in different forms at the Nord Pool financial market). A report written by Copenhagen Economics (2002) on the relevant power markets in the Nordic area [10] reached the conclusion that the bilateral market, the spot market, the Elbas market and the regulating power market all belong to the same relevant product market. Copenhagen Economics put much emphasis on the fact that supply substitution is feasible and likely to be swift, such that prices on one of the markets cannot rise significantly without generators quickly moving capacity from the other three markets to the market with the higher price. The generator does not need to make any significant technological adjustment but only needs to adjust the bid schedules on the spot market and on the regulating power market. These conclusions were backed by analysis of price correlation between all the wholesale sub-markets except for the market for bilateral contracts. In this report we will base our analysis on a relevant product market that include all the wholesale sub markets. 2.4 The relevant geographic marketThe European Commission has traditionally defined the geographic wholesale market for electricity as national. The argument is that the lack of international transmission capacity and the lack of formalized tools for international exchange of electricity effectively prevent demand and supply substitution outside the range of the national transmission networks. An important characteristic of the power market is the lack of opportunity to store the product. Electricity must be consumed in the same instant it is produced. The substitutability between different time periods is also restricted. The lack of possibilities for consumers to store electricity and the limited substitution between time periods implies that the market must be distinguished by the time at which the electricity is delivered. In some instances where the time dimension matters it is natural to include it as a characteristic of the product (for instance new-grown potatoes). In the electricity market the time dimension is important because it influences the width of the geographical market. During different time periods there will be capacity constraints in the transmission network. Such capacity constraints limit the range of suppliers that can offer electricity in a particular region. This concept of a relevant geographic market varying over time is a special feature of the electricity market. In competition cases concerning abusive behaviour in other markets the relevant market needs to be stable over a period of time in order for the competition authority to establish dominant position. This is not possible in the electricity market. What is crucial in establishing a dominant position on a given market is the ability to affect prices. If the prices can be affected from hour to hour then this period of time is long enough to establish dominance. In the electricity market, prices can be affected by the strategic behaviour of a generator from one hour to the next. Congestions in the inter-Nordic transmission network divide the market geographically. Because of capacity constraints in the network grids (so-called bottlenecks), the structure of the power market can vary from one hour to the next. In periods where there are no binding bottlenecks the relevant geographic market is delineated to the Nordic region (Norway, Sweden, Finland, Denmark). Even if the Nordic power market is also connected to other geographic regions like Poland and Germany, there are no indications that the relevant geographic market is ever bigger than the Nordic region. Copenhagen Economics has, in the above-mentioned report, analysed partial price correlations between areas, which clearly reject a hypothesis that the Nordic area is part of the same market as Germany. When bottlenecks bind in the Nordic transmission grids the spot market is divided into two or more areas with different prices (so-called price areas). In these periods it is quite obvious that a hypothetical monopolist operating in an area into which imports are restricted, will be able to increase price without loosing demand to suppliers located in other areas. Thus, the SSNIP test will be satisfied for such a price area. It is important to note that a hypothetical monopolist will be able to exercise market power also in an area with surplus supply, i.e. an area from which electricity is exported to surrounding areas. Market power may be utilised by preventing the price from falling to a lower level. This means that whenever the wholesale market is divided into different price areas, each area represents a separate geographic market. Furthermore, a hypothetical monopolist in an area with surplus supply may raise the price to the level of the surrounding areas, and thereby level out price differences between price areas. This means that a price area may be a relevant geographic market even in periods where the capacity in the transmission grids is not fully utilized. It should also be emphasised that there might be relevant geographic markets within each price area. Capacity constraints within price areas are handled by means of counter-purchase. Power companies operating in these markets may be able to exert market power towards the system operator. How important are the transmission constraints that divide the Nordic region into smaller relevant geographic markets? The table below shows export and import openness [11] in the Nordic countries. Trade openness
Source: Copenhagen Economics (2002) The data indicates that all Nordic countries are well connected to each other, especially Denmark and Sweden where the import/export capacity is well above 20 per cent of the internal generating capacity. In Norway the same indicators is approximately 20 per cent and in Finland slightly above 10 per cent. However, not all of the capacity of the transmission lines is fully utilised. For instance, the average degree of utilization of the transmission line from Sweden to Denmark West was just below 50 per cent in 2001. This indicates that there must be scope for increasing the effective transmission capacity by other means than increasing the physical capacity. In those hours where no congestion constraints occur, wholesale prices are identical in all Nordic countries and the Nordic area is an integrated relevant market. When congestion does occur, the relevant geographic market is separated into several geographic markets depending on the exact location of the congestion constraints. Some of these geographic markets are directly observable on the spot market because the congestion constraint is handled by means of Elspot price areas (see 2.2 and 1.2.4), giving rise to different prices in different price areas. Other geographic markets are only visible at the regulating power market, because the congestion constraint is handled by counter trade, giving rise to identical prices on the spot market, but different prices on the regulating power market. The distribution of price areas within the Nordic area and the frequency with which they occur varies from year to year primarily due to variations in weather conditions. In wet years there will be congestion constraints in the transmission grids from Norway to Denmark and Sweden. For instance, in wet years (as the years 2000 and 2001) the two Danish price areas – dominated by expensive thermal generation – were isolated in more than 30 per cent of the time. In "normal" years the figure drop to approx. 5 per cent. Export from Denmark may be constrained when wind production is high or there are large imports from Denmark. Copenhagen Economics (2002) has listed all combinations of price areas on the Nordic spot market. From this list we have calculated the frequency of different constellations of Elspot price areas. We have only counted the number of constellations involving three price areas or less. The table below shows the most encountered constellations of price areas in 2001 and the percentage of time in which they occur. The figures vary between different years.
The integrated Nordic region is the most frequent price area. It occurred in 52% of the time in 2001. Note however, that in 2002 the percentage was 35% hour. Half of the time or more the relevant geographic market has been smaller than the Nordic region. Footnotes[3] The U.S. Department of Justice and Federal Trade Commission (1992) [4] The EU Commission (1997) [5] For a discussion of the two approaches, see chapter V of Werden, Gregory J. (1992). [6] Office of Fair Trading (2001). [7] Confer the discussion in subsection 2.2. [8] The Norwegian Competition Authority (2002) [9] Press release 26 March 2003. [10] Copenhagen Economics (2002) [11] Export (import) openness is calculated as the permissible export (import) capacity divided by the total generation capcity. Version 1.0 October 2003 • © Danish Competition Authority. |
